1. Technical Field
The present invention relates generally to object-oriented graphical user interface technology. More specifically, the present invention is directed to a method for traversing a hierarchy of GUI components taken from disparate GUI toolkits.
2. Description of the Related Art
The earliest interactive computers relied on tele-typewriter (TTY) or text terminals for interactive communication with a human operator. These early forms of human-computer interaction (HCI) allowed for only text- or character-based information exchange. Many computer software products today utilize a graphical user interface or GUI (typically pronounced like “gooey”). A GUI is visual means of human-computer interaction that utilizes pictures or other visual representations besides text or characters.
Most GUIs make use of visual controls that are displayed on the user's display and actuated by user input. Typical visual controls include, but are not limited to, buttons, text fields (for entering text), radio buttons, checkboxes, selection boxes, and menu bars. In a typical GUI, a pointing device, such as a mouse, is used to move a cursor around a display and actuate visual controls. GUIs usually also make use of static display components, such as labels and icons, which are intended to be displayed, but generally have no input function, per se. Sometimes these static display components may serve an input role, however, when they are moved around on the display relative to other features on the display (e.g., dragging an icon of a file to a trash can icon to delete a file, for example).
Many GUIs are what is known as a “windowing” interface, because they arrange information visually on a display in the form of panels or “windows” superimposed on a background called a “desktop.” In many systems, windows may be dragged to different locations on the display with a pointing device, enlarged, reduced, made to overlap with other windows. Typically, a window will contain a number of visual controls to allow a user to interact with a computer program by actuating the controls in the window. A special form of window, known as a “dialog box,” is displayed by a program when some input is required from a user.
Windows, visual controls, and static display components are what are known as GUI components, because they are the building blocks that make up the GUI. Some GUI components, such as windows, are known as “container components” (or simply “containers”), because they may contain other components. For example, a window may contain visual controls, such as a button or menu bar, and static display components, such as text labels or icons. A container may also contain another container. For example, in some windowing-based word processors, the word processor itself occupies a (main) window, while each file under editing occupies another window within the main window.
Container components include windows, but may also include other components, which may be visible or invisible. For example, the JAVA™ programming language produced by Sun Microsystems, Inc. of Mountain View, Calif., defines various visible container components, such as windows and dialog boxes, as well as invisible container components, such as the “java.awt.Panel” container component, which is used solely to group a number of contained components into a single unit. Some examples of containers include, but are not limited to, windows, dialog boxes, panels, tabbed panels, notebook pages, and any other GUI components that have a capability of containing one or more other GUI components.
The actual functionality for providing basic operations on GUI components, such as displaying the components or detecting user input directed at the components (e.g., from pointing at or clicking on a component with a pointing device), is often provided by system-level software, such as an operating system. Generally speaking, applications will issue calls to system-level software for creating and maintaining GUIs, while the system-level software detects user input events that are directed at particular GUI components and sends event notifications to the applications that are responsible for those GUI components.
For example, the WINDOWS® operating system produced by Microsoft, Inc. of Redmond, Wash. provides services for the creation of GUIs and relaying of user input events to appropriate applications. The main interface for the WINDOWS® operating system itself is a GUI as well. In other settings, higher-level system software may operate on top of an operating system kernel (e.g., as a daemon or background process) to provide GUI services. For example, “X11” is an open-source GUI engine that operates as a process in an operating system. X11 adopts a client-server model in that an X11 server process accepts requests from applications (clients) for providing GUI services and relays user input events that pertain to particular GUI components to the applications associated with those components.
Alternatively, an application may contain its own code for providing GUI services. Typically, this code will come in the form of a reusable code library for performing basic GUI operations.
Many modern programming language implementations have built-in features for producing GUIs, usually either by providing an interface to GUI services provided by system-level software or by including libraries of low-level GUI code for which an interface in the programming language is provided. The JAVA™ programming language, for example, is an object-oriented programming language that includes standard application programming interfaces (APIs) for defining GUIs. Two APIs that are currently part of the JAVA™ programming language standard are the Abstract Windowing Toolkit (AWT) API and the Swing API (which is built on the AWT API). In the JAVA™ programming language, as is typical of object-oriented GUI APIs, each type of GUI component is defined as a class.
In an object-oriented programming language, a class is a definition of a data type that includes a collection of data, called member variables, and a set of operations that may be performed on the data, called methods (or alternatively, member functions). An actual collection of data in the data type defined by a class is called an object. In object-oriented programming (OOP) parlance, an object is said to be an “instance” of the class, because it is a data structure that is defined in accordance with the class. The run-time process of generating an object in an object-oriented programming language is called “instantiation,” and an object that exists at run-time is said to be “instantiated.”
Object-oriented programming languages also typically provide for what is known as “inheritance.” Using an inheritance a new class (called a “descendant” class) can be defined in terms of one or more existing classes (called “base” classes) so that the descendant class inherits one or more of the member variables or methods of the base class. For example, in the JAVA™ programming language's AWT API, “Container” is a descendant class of a base class called “Component,” the “Container” class will include at least some of the methods and member variables of “Container.” We thus say that “Container” is descended from “Component.” In many cases, a descendant class will include additional methods or member variables that are not inherited from the base class.
Also, a descendent class may be written so as to override the base class's code for a particular method. For example, the base class “Container” may have a method called “show,” for displaying a GUI component, which the descendant class “Container” inherits. Since displaying a container (which may contain other components) is more specific than displaying a generic GUI component, the “Container” class may define different code for “show” than that of the “Component” class.
This is important, since in most object-oriented languages, an object in a descendant class is treated as being a more specific instance of the base class. Thus, a “Container” object may be stored in a variable of type “Component,” or a method that takes a “Component” as an argument can also take a “Container” as an argument, since a “Container” will inherit characteristics (i.e., member variables and methods) from “Component.” This ability to treat objects from descendant classes as if they were instances of base classes is called “polymorphism.”
In an object-oriented GUI API, such as those provided by the JAVA™ programming language, GUI components are instantiated as objects, and relationships are established between the instantiated objects in order to define the placement and behavior of GUI components with respect to each other. For example, a “containment relation” is a relationship between GUI components that relates a container component to the components contained by that container component. In the JAVA™ programming language, for example, a component typically enters into a containment relation with a container through a method of the container called “add.”
A typical GUI component has one or more attributes that define particular properties of the component. For example, a “button” component in a typical windowing GUI will have attributes that define the size of the button on the display, the text or graphics displayed on the face of the button, the background color of the button, a keyboard shortcut associated with the button, and the like. In general, the portion of program code (e.g., function, method, subroutine, procedure, etc.) that instantiates a GUI component will also contain a number of lines of code that set the attributes for that component to desired values. In the JAVA™ programming language and other object-oriented programming systems, for example, components generally have methods that can be executed to set particular attributes of the component.
“Eclipse” is an open-source software platform and API, written in JAVA, that is designed to facilitate the creation of integrated development environments (IDEs). IDEs typically combine a text editor with various software development tools, such as compilers, linkers, debuggers, profilers, and the like through a unified user interface. The Eclipse platform, not surprisingly, provides its own support for creating GUIs. The Eclipse API includes a GUI toolkit called the “Standard Widget Toolkit,” or “SWT” for short. SWT has a similar object-oriented structure to AWT.
It is possible to use components from both the AWT and SWT toolkits in a single GUI. The Eclipse API includes a class called “SWT_AWT,” which includes methods that allow one to create a child component/container in one toolkit for association with a parent component/container in the other toolkit. The method “SWT_AWT.new_Frame,” for instance, allows one to create a “Frame” container in the JAVA™ AWT that is associated with a parent container (called a “Composite”) in the SWT. Likewise, “SWT_AWT.new_Shell” allows one to create a “Shell” container in the SWT that is associated with a parent “Container” (intentionally capitalized) in the JAVA™ AWT. Eclipse does not, however, provide a way to take a container in one toolkit and access its children that are in the other toolkit. This makes it impossible to fully traverse a GUI component tree that contains components/containers from both toolkits.
What is needed, therefore, is method and system to allow a GUI component tree containing components/containers from multiple toolkits to be traversed. The present invention provides a solution to these and other problems, and offers other advantages over previous solutions.